Anda di halaman 1dari 4


Serum Vitamin D and the Risk of Parkinson Disease

Paul Knekt, DPH; Annamari Kilkkinen, PhD; Harri Rissanen, MSc; Jukka Marniemi, PhD;
Katri Sääksjärvi, MSc; Markku Heliövaara, PhD

Objective: To investigate whether serum vitamin D level Main Outcome Measure: Parkinson disease inci-
predicts the risk of Parkinson disease. dence.

Design: Cohort study. Results: Individuals with higher serum vitamin D con-
centrations showed a reduced risk of Parkinson disease.
Setting: The study was based on the Mini-Finland The relative risk between the highest and lowest quar-
Health Survey, which was conducted from 1978 to tiles was 0.33 (95% confidence interval, 0.14-0.80) after
1980, with Parkinson disease occurrence follow-up adjustment for sex, age, marital status, education, alco-
through the end of 2007. During the 29-year fol- hol consumption, leisure-time physical activity, smok-
low-up period, 50 incident Parkinson disease cases ing, body mass index, and month of blood draw.
occurred. Serum 25-hydroxyvitamin D level was deter-
mined from frozen samples stored at baseline. Esti- Conclusions: The results are consistent with the sug-
mates of the relationship between serum vitamin D gestion that high vitamin D status provides protection
concentration and Parkinson disease incidence were
against Parkinson disease. It cannot, however, be ex-
calculated using the Cox model.
cluded that the finding is due to residual confounding
Participants: Three thousand one hundred seventy-
and further studies are thus needed.
three men and women, aged 50 to 79 years and free of
Parkinson disease at baseline. Arch Neurol. 2010;67(7):808-811.

ITAMIN D PLAYS AN IMPOR- logical data indicate that vitamin D defi-
tant role in the pathogen- ciency may contribute to its develop-
esis of skeletal disorders ment.5 The present cohort study
and calcium homeosta- investigated whether serum 25-hydroxyvi-
sis. 1 Vitamin D inad- tamin D level predicts Parkinson disease
equacy also predicts increased risk of other incidence in a population from northern
chronic conditions, eg, cancer,2 cardio- latitudes where exposure to the sun is lim-
vascular diseases,3 and type 2 diabetes ited and therefore vitamin D status is con-
mellitus. 4 Recently, chronically inad- tinuously low.
equate vitamin D intake was proposed to
play a significant role in the pathogenesis
of Parkinson disease.5 According to the
For editorial comment
suggested biological mechanism, Parkin- see page 795
son disease may be caused by a continu-
ously inadequate vitamin D status lead-
ing to a chronic loss of dopaminergic METHODS
neurons in the brain. The epidemiologi-
cal evidence of an association between vi-
tamin D and Parkinson disease is, how- The Mini-Finland Health Survey, carried out
ever, limited to cross-sectional studies6-8 from 1978 to 1980 in 40 areas of Finland, was
showing lower vitamin D status in pa- based on a 2-stage cluster sample (n=3637 men
and n=4363 women) drawn from the popu-
tients with Parkinson disease compared lation register to represent Finnish adults 30
with healthy controls. years and older.9 A total of 7217 individuals
Parkinson disease is a major cause of (90% of the sample) participated in the sur-
Author Affiliations: National disability in elderly individuals. Its risk fac- vey. Of these, 3173 individuals, aged 50 to
Institute for Health and tors are relatively unknown. However, 79 years, free of Parkinson disease and not
Welfare, Helsinki, Finland. both biological plausibility and epidemio- using antipsychotic medication to treat psy-



©2010 American Medical Association. All rights reserved.

Downloaded From: by David Perlmutter on 08/19/2012
chotic disorders (International Statistical Classification of Dis-
eases, 10th Revision codes F20-F39), were included in the Table 1. Selected Sex- and Age-Adjusted Baseline
present study. Characteristics by Parkinson Disease
Information on socioeconomic background, diseases, medi-
cations, and lifestyle was collected via questionnaires and in- Parkinson Disease, %
terviews.9 At the baseline examinations, height and weight were
Noncases Cases P for
measured, and the body mass index was calculated as weight
(n = 3123) (n = 50) Heterogeneity
in kilograms divided by height in meters squared. Casual blood
pressure was measured with the auscultatory method and hy- Age, a y, mean (SD) 61.8 (8.0) 60.4 (6.5) .23
pertension was defined as systolic blood pressure 160 mm Hg Male b 43.1 47.2 .56
or higher and diastolic blood pressure 95 mm Hg or higher or Summer season, 18.3 12.7 .25
the use of antihypertensive medication. Blood samples were
More than basic 19.8 23.2 .55
taken and the cholesterol concentrations determined by an au- education
toanalyzer modification (Auto-Analyzer Methodology N-24a Married 65.9 68.7 .66
and N-77; Technicon, Tarrytown, New York) of the Liebermann- Regular leisure-time 11.3 7.6 .42
Burchard reaction. The serum samples were kept frozen at −20°C physical activity
until 2002, when serum 25-hydroxyvitamin D concentrations Smoker 18.7 6.2 .02
were determined using radioimmunoassay (DiaSorin, Stillwa- Alcohol consumption, 29.6 (86.2) 13.1 (51.1) .15
ter, Minnesota). The interassay coefficient of variation of 25- grams of ethanol/wk,
hydroxyvitamin D concentration determination was 7.8% at the mean (SD)
mean level of 47.3 nmol/L (n = 167). The intra-assay coeffi- Hypertension 35.7 21.1 .03
cient of variation was 6.4%. The samples were run as single Body mass index, c mean 26.8 (4.2) 26.5 (3.3) .62
samples. The right assay level was ensured by using the refer- (SD)
Diabetes mellitus 8.7 2.7 .13
ence serum validated by the National Institute of Standards and
Serum total cholesterol 283.4 (52.9) 279.9 (51.7) .65
Technology (Standard Reference Material 968c Fat-Soluble Vi-
level, mg/dL, mean
tamins; National Institute of Standards and Technology, Gaith- (SD)
ersburg, Maryland). The laboratory also participates in the ex- Serum 41.8 (19.5) 36.3 (18.5) .05
ternal quality control program run by Labquality Oy (Helsinki, 25-hydroxyvitamin D
Finland). In addition, the laboratory’s vitamin D concentra- level, nmol/L, mean
tion measurement method is accredited by the Finnish Ac- (SD)
creditation Service (FINAS, T077).
Parkinson disease cases (International Statistical Classifica- SI conversion factor: To convert total cholesterol to millimoles per liter,
tion of Diseases, 10th Revision code G20) were identified through multiply by 0.0259.
a Adjusted for sex.
linkage with the nationwide Drug Imbursement Register of the b Adjusted for age.
Social Insurance Institution, using individual social security c Calculated as weight in kilograms divided by height in meters squared.
codes as the identity link. All individuals in Finland with Par-
kinson disease are eligible for medication free of charge. To ob-
tain this allowance, the patient must apply for it and attach a
certificate written by the treating neurologist stating that all the tal status, education, alcohol consumption, leisure-time physi-
diagnostic criteria for Parkinson disease are met. This certifi- cal activity, smoking status, body mass index, and month of
cate must include symptom history and reports of clinical find- blood draw. In a third model, the eventual intermediary vari-
ings, including the presence of resting tremor, bradykinesia, ables hypertension and total serum cholesterol level were fur-
and/or muscle rigidity. A Social Insurance Institution neurolo- ther included. Potential effect modification of sex, age, sea-
gist must agree with the diagnosis as described on the certifi- son, hypertension, body mass index, and serum cholesterol level
cate for medication costs to be reimbursed. In an ongoing vali- on the association between vitamin D level and Parkinson dis-
dation of the register, the certificates for Parkinson disease drug ease incidence was studied by including interaction terms in
reimbursement and selected hospital records were reevalu- the second model. All analyses were carried out using SAS soft-
ated retrospectively by a neurologist according to the Na- ware version 9 (SAS Institute Inc, Cary, North Carolina).
tional Institute of Neurological Disorders and Stroke diagnos-
tic criteria for Parkinson disease.10,11 Of the originally identified RESULTS
Parkinson disease cases reviewed, 80% met criteria for Parkin-
son disease ( J. Lyytinen, MD, PhD, oral communication, Janu-
ary 2009), consistent with other estimates of the percentage of At baseline, Parkinson disease cases more often were non-
people clinically diagnosed with parkinsonism in a general popu- smokers and did not have hypertension or diabetes than
lation who meet strict Parkinson disease criteria.12 The fol- subjects who were free of the disease (Table 1). Serum
low-up time was defined as the number of days from the base- vitamin D concentration was lower among Parkinson dis-
line examination to the dates of Parkinson disease occurrence, ease cases and it was also associated with age, sex, mari-
death, or end of follow-up, whichever came first. During a 29- tal status, education, leisure-time physical activity, smok-
year follow-up from 1978 to 2007, 50 Parkinson disease cases ing, alcohol consumption, body mass index, diabetes,
were identified. hypertension, serum cholesterol level, and the season of
The Cox proportional hazards model was used to estimate measurement (Table 2).
the strength of association between serum vitamin D level and
Parkinson disease incidence as relative risks (RRs) and their
A significant inverse association between sex- and age-
95% confidence intervals (CIs) between quartiles of serum vi- adjusted serum vitamin D level and Parkinson disease in-
tamin D level.13 Test for trend was based on the likelihood ra- cidence was found (Table 3). The RR of the disease be-
tio test by including serum vitamin D level as a continuous vari- tween the highest and lowest quartiles of vitamin D
able in the 3 models. The first model included age and sex as concentration was 0.35 (95% CI, 0.15-0.81; P for
potential confounders. The second model further included mari- trend=.006). After further adjustment for the potential



©2010 American Medical Association. All rights reserved.

Downloaded From: by David Perlmutter on 08/19/2012
Table 2. Selected Sex- and Age-Adjusted Baseline Characteristics by Serum 25-Hydroxyvitamin D Level Quartiles a

Serum 25-Hydroxyvitamin D Level, %

Quartile 1 Quartile 2 Quartile 3 Quartile 4

(n = 774) (n = 819) (n = 778) (n = 792) P for Trend
Age, b y, mean (SD) 63.8 (8.1) 62.2 (8.1) 61.2 (7.8) 60.0 (7.6) ⬍.001
Male c 45.1 43.4 42.6 41.8 .18
Summer season, Jun-Sep 2.7 12.7 21.1 36.1 ⬍.001
More than basic education 12.7 18.3 24.6 23.7 ⬍.001
Married 61.9 65.6 65.8 70.5 ⬍.001
Regular leisure-time physical activity 5.5 10.3 12.9 15.9 ⬍.001
Smoker 22.3 17.3 18.8 15.6 .002
Alcohol consumption, grams of ethanol/wk, mean (SD) 22.0 (71.3) 25.9 (72.1) 32.0 (96.1) 37.6 (98.8) ⬍.001
Hypertension 37.2 36.6 35.5 32.4 .04
Body mass index, d mean (SD) 26.7 (4.6) 27.1 (4.3) 27.1 (4.0) 26.2 (3.6) .03
Diabetes mellitus 11.2 8.9 8.2 6.1 ⬍.001
Serum total cholesterol level, mg/dL, mean (SD) 277.6 (54.4) 281.9 (52.1) 285.3 (52.9) 288.8 (51.0) ⬍.001

SI conversion factor: To convert total cholesterol to millimoles per liter, multiply by 0.0259.
a Quartiles for men: 8 to 28, 29 to 41, 42 to 56, and 57 to 159 nmol/L; for women: 7 to 25, 26 to 36, 37 to 49, and 50 to 151 nmol/L.
b Adjusted for sex.
c Adjusted for age.
d Calculated as weight in kilograms divided by height in meters squared.

Table 3. RRs With 95% CIs for Parkinson Disease Cases by Baseline Serum 25-Hydroxyvitamin D Level a

Serum 25-Hydroxyvitamin D Level, RR (95% CI)

Quartile 1 Quartile 2 Quartile 3 Quartile 4 P for Trend

No. of Parkinson disease cases 17 15 10 8
Sex- and age-adjusted model 1[Reference] 0.73 (0.36-1.46) 0.47 (0.21-1.03) 0.35 (0.15-0.81) .006
Multivariate model A b 1[Reference] 0.72 (0.36-1.46) 0.48 (0.22-1.08) 0.33 (0.14-0.80) .006
Multivariate model B c 1[Reference] 0.72 (0.36-1.45) 0.48 (0.21-1.07) 0.33 (0.14-0.78) .005

Abbreviations: CI, confidence interval; RR, relative risk.

a Quartiles for men: 8 to 28, 29 to 41, 42 to 56, and 57 to 159 nmol/L; for women: 7 to 25, 26 to 36, 37 to 49, and 50 to 151 nmol/L.
b Model A further included marital status (married or other), education (basic or intermediate/high), alcohol consumption (0, ⬍5, or ⱖ5 g/d), leisure-time
physical activity (no/light or heavy), smoking status (none or current), body mass index, and month of blood draw.
c Model B further included hypertension and serum cholesterol level.

confounders, including body mass index, leisure-time Vitamin D is obtained from diet and is photosynthe-
physical activity, smoking, education, marital status, al- sized in the skin by the action of solar UV-B radiation.
cohol consumption, and month of blood draw, the as- This study was carried out in Finland, an area with re-
sociation persisted (RR, 0.33; 95% CI, 0.14-0.80; P for stricted sunlight exposure, and is thus based on a popu-
trend=.006). Further adjustment for serum cholesterol lation with a continuously low vitamin D status. Accord-
level and hypertension or exclusion of the disease cases ingly, the mean serum vitamin D level in the present
occurring during the first 2 years of follow-up did not population was about 50% of the suggested optimal level
notably alter the results either. Inclusion of an interac- (75-80 nmol/L).14 Our findings are thus consistent with
tion term between vitamin D level and sex, age, body mass the hypothesis5 that chronic inadequacy of vitamin D is
index, serum cholesterol level, blood pressure, and the a risk factor for Parkinson disease.
season of measurement did not notably alter the results As far as we know, this is the first longitudinal study
(data not shown). to investigate the association between vitamin D status
and subsequent Parkinson disease occurrence. In line with
COMMENT our finding, however, previous cross-sectional studies
demonstrated the higher prevalence of hypovitaminosis
This cohort study shows that low serum vitamin D level D in patients with Parkinson disease than in healthy con-
predicts an elevated risk of Parkinson disease inci- trols.6-8 The exact mechanisms by which vitamin D may
dence. Individuals with a serum vitamin D concentra- protect against Parkinson disease are not fully under-
tion of at least 50 nmol/L had a 65% lower risk than those stood. Vitamin D has, however, been shown to exhibit
with values less than 25 nmol/L after adjustment for sev- neuroprotective effects through antioxidative mecha-
eral potential confounders. Despite the overall low vita- nisms, neuronal calcium regulation, immunomodula-
min D levels in the study population, a dose-response re- tion, enhanced nerve conduction, and detoxification
lationship was also found. mechanisms.5,15,16



©2010 American Medical Association. All rights reserved.

Downloaded From: by David Perlmutter on 08/19/2012
The vitamin D receptors and an enzyme responsible Financial Disclosure: None reported.
for the formation of the active form 1,25-hydroxyvita- Funding/Support: This work was supported by Na-
min D have been found in high levels in the substantia tional Institutes of Health grant NIH/NIEHS R01
nigra, the region of the brain affected most by Parkin- ES012667.
son disease.15 This raises the possibility that chronic in- Role of the Sponsor: The National Institutes of Health
adequacy of vitamin D leads to the loss of dopaminergic did not participate in the design or conduct of the study;
neurons in the substantia nigra region and further Par- the collection, analysis, or interpretation of the data; or
kinson disease. the preparation, review, or approval of the manuscript.
The strengths of the present study are the apparent long-
term inadequacy of vitamin D17 and the prospective
design. There are, however, some weaknesses. First, the REFERENCES
small number of cases may have caused instable results.
Second, only a single measurement of serum 25- 1. Grant WB. Epidemiology of disease risks in relation to vitamin D insufficiency.
hydroxyvitamin D was available, which fails to take into Prog Biophys Mol Biol. 2006;92(1):65-79.
account the intraindividual seasonal variation. However, 2. Kilkkinen A, Knekt P, Heliovaara M, et al. Vitamin D status and the risk of lung
no measurements were carried out during July and no in- cancer: a cohort study in Finland. Cancer Epidemiol Biomarkers Prev. 2008;
teraction between serum vitamin D concentration and sea-
3. Giovannucci E, Liu Y, Hollis BW, Rimm EB. 25-Hydroxyvitamin D and risk of myo-
son (sunny vs dark period) was observed. Serum vitamin cardial infarction in men: a prospective study. Arch Intern Med. 2008;168(11):
D level is relatively stable over time.18 The possibility that 1174-1180.
levels changed during long-term storage cannot, how- 4. Knekt P, Laaksonen M, Mattila C, et al. Serum vitamin D and subsequent occur-
ever, be excluded.19,20 Third, it is possible that the study rence of type 2 diabetes. Epidemiology. 2008;19(5):666-671.
5. Newmark HL, Newmark J. Vitamin D and Parkinson’s disease—a hypothesis.
population includes undefined Parkinson disease cases and Mov Disord. 2007;22(4):461-468.
also that all patients with Parkinson disease diagnoses are 6. Evatt ML, Delong MR, Khazai N, Rosen A, Triche S, Tangpricha V. Prevalence of
not definite cases. Because of the low prevalence of the dis- vitamin D insufficiency in patients with Parkinson disease and Alzheimer disease.
ease, the former error is not of great importance. The lat- Arch Neurol. 2008;65(10):1348-1352.
7. Sato Y, Honda Y, Iwamoto J, Kanoko T, Satoh K. Abnormal bone and calcium
ter error may, however, have biased the estimates of the
metabolism in immobilized Parkinson’s disease patients. Mov Disord. 2005;
strength of association. Fourth, the limited information on 20(12):1598-1603.
dietary intake of vitamin D is of potential concern. The ma- 8. Sato Y, Kikuyama M, Oizumi K. High prevalence of vitamin D deficiency and re-
jor dietary source of vitamin D is fatty fish, whose con- duced bone mass in Parkinson’s disease. Neurology. 1997;49(5):1273-1278.
sumption is also suggested to be beneficial against Parkin- 9. Aromaa A, Heliövaara M, Impivaara O, Knekt P, Maatela J. The Execution of the
Mini-Finland Health Survey. Part 1: Aims, Methods, and Study Population [in Finn-
son disease because of ␻-3 polyunsaturated fatty acids.21 ish with English summary]. Helsinki, Finland: Publications of the Social Insur-
The findings are, however, contradictory,22 and vitamin D ance Institution; 1989. Publication ML:88.
level has several other determinants.23 Fifth, the risk fac- 10. Gelb DJ, Oliver E, Gilman S. Diagnostic criteria for Parkinson disease. Arch Neurol.
tors for Parkinson disease are not well known, and there- 1999;56(1):33-39.
11. Jankovic J. Parkinson’s disease: clinical features and diagnosis. J Neurol Neu-
fore, despite comprehensive adjustments for potential con-
rosurg Psychiatry. 2008;79(4):368-376.
founders, residual confounding may still remain. 12. Schrag A, Ben-Shlomo Y, Quinn N. How valid is the clinical diagnosis of Parkin-
In conclusion, our results are in line with the hypoth- son’s disease in the community? J Neurol Neurosurg Psychiatry. 2002;73(5):
esis that low vitamin D status predicts the development 529-534.
of Parkinson disease. Because of the small number of cases 13. Cox DR. Regression models and life-tables. J R Stat Soc Series B. 1972;34:187-220.
14. Bischoff-Ferrari HA, Giovannucci E, Willett WC, Dietrich T, Dawson-Hughes B.
and the possibility of residual confounding, large co- Estimation of optimal serum concentrations of 25-hydroxyvitamin D for mul-
hort studies are needed. In intervention trials focusing tiple health outcomes. Am J Clin Nutr. 2006;84(1):18-28.
on effects of vitamin D supplements, the incidence of Par- 15. Eyles DW, Smith S, Kinobe R, Hewison M, McGrath JJ. Distribution of the vita-
kinson disease merits follow up. min D receptor and 1 alpha-hydroxylase in human brain. J Chem Neuroanat. 2005;
16. Buell JS, Dawson-Hughes B. Vitamin D and neurocognitive dysfunction: pre-
Accepted for Publication: September 8, 2009. venting “D”ecline? Mol Aspects Med. 2008;29(6):415-422.
Correspondence: Paul Knekt, DPH, National Institute 17. Huotari A, Herzig KH. Vitamin D and living in northern latitudes—an endemic risk
for Health and Welfare, PL 30, 00271 Helsinki, Finland area for vitamin D deficiency. Int J Circumpolar Health. 2008;67(2-3):164-178.
( 18. Platz EA, Leitzmann MF, Hollis BW, Willett WC, Giovannucci E. Plasma 1,25-
dihydroxy- and 25-hydroxyvitamin D and subsequent risk of prostate cancer. Can-
Author Contributions: Prof Knekt had full access to all cer Causes Control. 2004;15(3):255-265.
of the data in the study and takes responsibility for the 19. Norris RL, Thomas MJ, Craswell PW. Assessment of a two-step high-performance
integrity of the data and the accuracy of the data analy- liquid chromatographic assay using dual-wavelength ultraviolet monitoring for 25-
sis. Study concept and design: Knekt and Kilkkinen. Ac- hydroxyergocalciferol and 25-hydroxycholecalciferol in human serum or plasma.
J Chromatogr. 1986;381(1):53-61.
quisition of data: Knekt, Marniemi, and Heliövaara. Analy-
20. Stamp TC, Round JM. Seasonal changes in human plasma levels of 25-
sis and interpretation of data: Knekt, Kilkkinen, Rissanen, hydroxyvitamin D. Nature. 1974;247(442):563-565.
and Sääksjärvi. Drafting of the manuscript: Knekt. Criti- 21. Gao X, Chen H, Fung TT, et al. Prospective study of dietary pattern and risk of
cal revision of the manuscript for important intellectual con- Parkinson disease. Am J Clin Nutr. 2007;86(5):1486-1494.
tent: Knekt, Kilkkinen, Rissanen, Marniemi, Sääksjärvi, 22. Chen H, Zhang SM, Hernan MA, Willett WC, Ascherio A. Dietary intakes of fat
and risk of Parkinson’s disease. Am J Epidemiol. 2003;157(11):1007-1014.
and Heliövaara. Statistical analysis: Knekt. Obtained fund- 23. van Dam RM, Snijder MB, Dekker JM, et al. Potentially modifiable determinants
ing: Knekt. Administrative, technical, and material sup- of vitamin D status in an older population in the Netherlands: the Hoorn Study.
port: Kilkkinen. Study supervision: Knekt. Am J Clin Nutr. 2007;85(3):755-761.



©2010 American Medical Association. All rights reserved.

Downloaded From: by David Perlmutter on 08/19/2012